The Internet Tidal Wave

The situation

In the spring of 1995 Microsoft was at the height of its power. MS-DOS and Windows ran on most of the world’s personal computers, Office was becoming the standard for documents and spreadsheets, and Windows 95, due that August, was the most anticipated software release in history. Microsoft’s plans for online services centred on MSN, a proprietary network in the style of AOL and CompuServe, to be bundled with Windows 95.

The World Wide Web was a few years old. Netscape, founded in 1994, had released Navigator, and by the summer of 1995 some accounts put its share of web browsing above 80 percent. Netscape planned to go public in August. Its leaders talked openly about the browser becoming the place where applications ran, which would make the operating system underneath less important.

Gates had written about the internet before, but in May 1995 he sent a memo that changed its rank. It was addressed to the executive staff and their direct reports.

What the memo argues

The memo opens with a comparison that tells readers how serious the author is: “The Internet is the most important single development to come along since the IBM PC was introduced in 1981.” Then it assigns priority in plain terms: “Now I assign the Internet the highest level of importance.” Gates adds that the internet is “even more important than the arrival of the graphical user interface,” the shift that had made Windows itself.

The core of the argument is a threat analysis. Gates names the competitor: “A new competitor ‘born’ on the Internet is Netscape.” And he names the mechanism by which it could hurt Microsoft. Netscape, he writes, is pursuing a strategy to “move the key API into the client to commoditize the underlying operating system.” In other words, if programs were written for the browser rather than for Windows, any operating system that could run the browser would do, and Windows would lose the advantage that made it valuable.

The memo then turns practical. Gates asks every executive to use the internet themselves, pointing them to sites like Yahoo that catalogued it. He describes how products across the company, from Office to MSN to help files, should be rebuilt around web formats, and he assigns leaders: “Paul Maritz will lead the Platform group,” with others responsible for applications and content. On MSN he is blunt about what the internet means for the proprietary network: the merger of online and internet businesses “creates a major challenge for MSN.”

It closes with a short line that summarises the whole: “The Internet is a tidal wave. It changes the rules.”

The decision it carried

The memo did not announce a single product. Its decision was a reallocation of attention: the internet would become the organising priority of the company, above the products that were making Microsoft’s money. Practical choices followed in the next two years. Microsoft licensed browser technology and released Internet Explorer in August 1995. It gave the browser away free. It shifted MSN from a closed network toward the web. And, most consequentially, it built Internet Explorer into Windows itself, so that every new PC came with a browser already installed.

The alternatives were real. Microsoft could have treated the web as one channel among many and continued to build MSN as a walled garden. It could have competed with Netscape purely on product quality. Or it could have tried to buy or partner with Netscape. The memo implicitly rejected the first; the company tried a version of the second and leaned heavily on distribution through Windows.

What happened

The turn worked faster than most observers expected. Internet Explorer improved with each release, and Internet Explorer 4, in October 1997, was competitive with Navigator. Because it came with Windows and cost nothing, its share grew rapidly. By the end of the decade Netscape had lost most of its market, and AOL acquired the company.

The methods used had a cost. In 1998 the US Justice Department and a group of states sued Microsoft, arguing among other things that it had illegally maintained its operating system monopoly by tying Internet Explorer to Windows and restricting what PC makers could do. The memo was part of the evidence, since it showed that Microsoft understood the browser as a threat to Windows. A trial court ruled against Microsoft in 2000 and ordered the company broken up; in 2001 an appeals court upheld the finding of illegal monopoly maintenance but reversed the break-up, and the case ended in a settlement with conduct restrictions.

The deeper outcome is that the memo defended Windows successfully against the browser, but the web’s most valuable business, search and advertising, was built by companies that did not exist in 1995. The memo mentions search sites as places executives should visit, not as a business Microsoft should lead.

How to read it as a product manager

The memo is worth studying for its structure as much as its content.

It starts with a calibration. By comparing the internet to the IBM PC and the graphical interface, the two events that made Microsoft, Gates told readers how large a change he meant, in terms they could not misread.

It names a specific competitor and a specific mechanism. Vague warnings about “disruption” rarely change behaviour. “Netscape could make Windows interchangeable by moving the important interfaces into the browser” is a hypothesis that teams can act against.

It ranks priorities explicitly. “Highest level of importance” is a decision about trade-offs: when the internet conflicts with something else, the internet wins. Most strategy documents avoid saying this because it creates losers.

It assigns owners. The named assignments are what turned the memo into a reorganisation.

It also shows a limit that is easy to miss. The memo is written from the point of view of the existing business: what does the internet do to Windows, Office and MSN? It is much less clear about what customers would want from the web that Microsoft could uniquely provide. A strategy built on defending the core can be correct and still leave the new market’s biggest opportunity to someone else.

What the memo missed, and what is contested

Two points are fair to raise. First, the memo frames the browser mainly as a threat to be neutralised, and the company’s later tactics against Netscape were found unlawful in court. Analysing a decision is different from judging people, but it is reasonable to say the strategy carried legal risk that the memo did not consider.

Second, historians debate how much the memo changed and how much it described changes already under way inside Microsoft, where several engineers had been arguing for an internet strategy. The memo’s fame may overstate a single document’s role in a company-wide shift. What it clearly did was make the shift official and give it the chief executive’s full authority.

The decision, examined closely

What triggered it

Two things changed in the year before the memo. Usage of the web was growing explosively, and Netscape had turned the browser from a viewer of documents into something that looked like a platform. Gates’s memo makes clear which of these worried him more. Growth alone was an opportunity; a platform that could run on any operating system was a threat to the economics of Windows.

The constraints

Microsoft’s calendar was dominated by Windows 95, the largest launch in its history, due three months later. Its online strategy was built around MSN, which could not simply be cancelled. And Netscape was weeks from a public offering that would give it capital and attention. The memo had to redirect a company already committed to a very different plan, without derailing the launch that paid for everything.

What the author was optimising for

Read closely, the memo optimises for keeping Windows central. Almost every assignment asks how a product can be made to work with the internet so that Microsoft’s platform stays the place where software is written. That is a coherent objective for the owner of Windows. It is also narrower than “win the internet,” and it explains why the strategy focused on the browser rather than on the services, such as search and advertising, that later became the web’s main business.

Second-order effects

Winning the browser war had consequences Microsoft did not intend. The antitrust case consumed senior attention for years and placed limits on the company’s conduct. Internet Explorer’s dominance removed the pressure to improve it, so it stagnated in the early 2000s, which opened the door for Firefox and later Chrome. And the memo’s framing, the browser as a threat to Windows, may have made it harder for Microsoft to see the browser as a platform worth leading in its own right.

The counterfactual

It is reasonable to ask what would have happened if Microsoft had competed on quality and kept the browser separate from Windows. It might have lost the browser market for longer, but avoided the lawsuit. Or Netscape’s vision of the browser as the platform might have arrived years earlier. No one can know. What can be said is that the memo’s diagnosis was correct, and that the choice of methods, more than the diagnosis, determined both the victory and its costs.

Sources and limitations

Quotations are taken from a full reprint of the memo checked against an excerpt published by the National Constitution Center; the original surfaced as evidence in the Justice Department’s antitrust case, and copies of the court exhibit are poor scans. Details of Netscape’s market share and the browser war come from historical accounts rather than audited data. The account of the antitrust case is a summary of widely reported outcomes. The interpretation of what the memo missed is the author’s.